Stability-Aware Predictive Clustering for 6G Integrated TN–NTN Cell-Free Massive MIMO
Year: 2026
Type: conference-paper
Abstract: The integration of terrestrial and non-terrestrial networks (TN-NTN) emerges as a cornerstone of 6G, promising global 3D coverage while introducing critical challenges in latency asymmetry, dynamic mobility and heterogeneous propagation conditions. This paper introduces P-STAR, a predictive stability-aware clustering framework for user-centric cell-free massive MIMO that fundamentally addresses the stability-efficiency trade-off in integrated TN-NTN systems. Our approach employs Kalman filtering for accurate channel prediction and incorporates a novel stability penalty mechanism to mitigate frequent handovers. We formulate the dynamic association problem as a long-term network utility maximization with latency-aware constraints, solved through an efficient three-phase heuristic algorithm with proven convergence guarantees. Extensive system-level simulations demonstrate that P-STAR achieves significant improvements in network utility, fairness, URLLC latency compliance, and cell-edge performance while substantially reducing handover rates compared to state-of-the-art baselines. These results, validated through rigorous Monte Carlo analysis, position P-STAR as an essential enabler for seamless and efficient 3D connectivity in next-generation networks.
Authors: Ainur Daurembekova, Hans D. Schotten
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166535224